Skip to contents

Length-Length Conversion Coefficients for Expanded Taxa

Source: R/fishbase.R
get_length_length_coeffs.Rd

Attaches length-length conversion coefficients from the FishBase / SeaLifeBase popll table to an expanded taxa table. These convert one length measurement into another via Length1 = aL + bL * Length2 — note the direction: Length2 is the predictor — which is required whenever a survey records a different length type from the one the length-weight coefficients are expressed in.

Usage

get_length_length_coeffs(
  expanded,
  length_types = c("TL", "FL"),
  version = "latest"
)

Arguments

expanded

A table as returned by expand_taxonomic_info(), optionally already restricted to an FAO area.

length_types

Character vector of length measurement codes to keep. Both Length1 and Length2 must be in this set. Defaults to c("TL", "FL") — total and fork length. Pass NULL to apply no filter.

version

FishBase / SeaLifeBase release to read. See resolve_db_version().

Value

A tibble with:

  • alpha3_code, original_name, SpecCode, species_found, server: carried through from expanded.

  • Length1, Length2: the measurement types converted from and to.

  • aL, bL: intercept and slope of the conversion.

  • LengthMin, LengthMax, Number, Sex, r2: study metadata.

Details

Note the naming: the popll table calls the conversion coefficients a and b, the same names the length-weight table uses for entirely different quantities. They are renamed to aL / bL here so that the two tables can be joined without collision.

The direction is Length1 = aL + bL * Length2, not the reverse. Releases up to 4.9.0 documented it backwards. Reading it the wrong way round inverts every ratio, which is worse than not converting at all. The data settles it: over FishBase 25.04, rows with Length1 = "TL", Length2 = "FL" have a median bL of 1.052 (n = 6,641) and Length1 = "TL", Length2 = "SL" a median of 1.204 (n = 8,848). Total length exceeds both fork and standard length, and standard length by more, so bL can only be the multiplier onto Length2.

This is load-bearing wherever a survey's recorded length type differs from the coefficient type. Timor-Leste's v1 survey form records fork length while its coefficients are predominantly total length, so without this conversion every v1 catch record carrying a measured length yields no weight at all.

See also

get_length_weight_coeffs(), get_taxa_morphometrics()

Examples

if (FALSE) { # \dontrun{
taxa <- data.frame(
  alpha3_code = "EMP",
  scientific_name = "Lethrinidae"
)
ll <- taxa |>
  expand_taxonomic_info() |>
  get_length_length_coeffs()
} # }